Probably you have seen this girl on the cover page of National Geographic magazine.She was photographed in 1984 aged 12 by National Geographic photographer Steve McCurry . After 18 years, Steve McCurry traced back the Afghan Girl who became popular (her photograph was compared to that of Mona Lisa) due to the fact that she represented the suffering of young girls in Afghanistan at the time. Steve McCurry who had a difficult time identifying her as many women came up and claimed that they were the "Afghan Girl". In this scenario, he took the help of John Daugman, pioneer of iris recognition algorithm which are being used today in most of the deployed commercial iris scanners.
How did John Daugman confirmed that identity of the Afghan Girl?
He confirmed her identity by using Iris Recognition algorithms on magnified images of eyes regions in the 1984 and 2002 photographs.He computed the Iris codes of both of her eyes of the photographs taken in 1984 and 2002. The image on right, shows the localization of iris and its boundaries along with the iris code. An iris code is computed by demodulation of the iris pattern. It uses complex valued 2D Gabor wavelets. The process of extracting the structure of the iris as a consequence of phasors is done in a doubly-dimensionless polar co-ordinate system that is never changing with reference to the iris and also to the diameter of pupil within in the iris.
The phase description represents iris pattern in only 256 bytes.Iris code is based on the failure of a test of statistical independence.The key point here is that IrisCode is statistically guaranteed to pass a test of independence when compared against the IrisCode of different eye, and it will fail when tested against the same eye from which the IrisCode was computed. Hence, the failure of a test of statistical independence is very important in iris recognition. After computing the IrisCode of both the eyes of the 1984 and 2002 photograph, John Daughman, using the matching algorithm on the IrisCode computed the Hamming Distance.Hamming distance is basically calculated by the number bits matched divided by the total number of bits compared, which resulted in 0.24 for her left eye and 0.31 for her right eye. When one looks at the histogram that arises when different irises are compared by their IrisCodes, theses measured HD’s are so far out on the distribution tail that it is impossible for two different eyes to show such little dissimilarity. Hence, by using iris recognition the Afghan Girl, Sharbut Gula was identified 18 years later .
Reference: Daughman, John. "How the Afghan Girl Was Identified by Her Iris Patterns." How the Afghan Girl Was Identified by Her Iris Patterns. N.p., n.d. Web. 13 Apr. 2014.SteveMcCurryStudios, dir. "Search For "The Afghan Girl" HQ (3/4)." Search For "The Afghan Girl" N.d. YouTube. Web. 15 Apr.The Afghan Girl.[Image].Retrieved from http://ngm.nationalgeographic.com/2002/04/afghan-girl/index-text
Red Book Unit 1
Who is She?Eyes Have It - Retina and Iris Identification
How the Afghan Girl was Identified by Her Iris Patterns
http://stevemccurry.com/galleries
Afghanistan
Steve McCurry in her eyes
Probably you have seen this girl on the cover page of National Geographic magazine.She was photographed in 1984 aged 12 by National Geographic photographer Steve McCurry . After 18 years, Steve McCurry traced back the Afghan Girl who became popular (her photograph was compared to that of Mona Lisa) due to the fact that she represented the suffering of young girls in Afghanistan at the time. Steve McCurry who had a difficult time identifying her as many women came up and claimed that they were the "Afghan Girl". In this scenario, he took the help of John Daugman, pioneer of iris recognition algorithm which are being used today in most of the deployed commercial iris scanners.
How did John Daugman confirmed that identity of the Afghan Girl?
He confirmed her identity by using Iris Recognition algorithms on magnified images of eyes regions in the 1984 and 2002 photographs.He computed the Iris codes of both of her eyes of the photographs taken in 1984 and 2002. The image on right, shows the localization of iris and its boundaries along with the iris code. An iris code is computed by demodulation of the iris pattern. It uses complex valued 2D Gabor wavelets. The process of extracting the structure of the iris as a consequence of phasors is done in a doubly-dimensionless polar co-ordinate system that is never changing with reference to the iris and also to the diameter of pupil within in the iris.
The phase description represents iris pattern in only 256 bytes.Iris code is based on the failure of a test of statistical independence.The key point here is that IrisCode is statistically guaranteed to pass a test of independence when compared against the IrisCode of different eye, and it will fail when tested against the same eye from which the IrisCode was computed. Hence, the failure of a test of statistical independence is very important in iris recognition. After computing the IrisCode of both the eyes of the 1984 and 2002 photograph, John Daughman, using the matching algorithm on the IrisCode computed the Hamming Distance.Hamming distance is basically calculated by the number bits matched divided by the total number of bits compared, which resulted in 0.24 for her left eye and 0.31 for her right eye. When one looks at the histogram that arises when different irises are compared by their IrisCodes, theses measured HD’s are so far out on the distribution tail that it is impossible for two different eyes to show such little dissimilarity. Hence, by using iris recognition the Afghan Girl, Sharbut Gula was identified 18 years later .
Reference:
Daughman, John. "How the Afghan Girl Was Identified by Her Iris Patterns." How the Afghan Girl Was Identified by Her Iris Patterns. N.p., n.d. Web. 13 Apr. 2014.SteveMcCurryStudios, dir. "Search For "The Afghan Girl" HQ (3/4)." Search For "The Afghan Girl" N.d. YouTube. Web. 15 Apr.The Afghan Girl.[Image].Retrieved from http://ngm.nationalgeographic.com/2002/04/afghan-girl/index-text